摘要: This paper presents the results of the development and validation of a tool created at the LARCASE laboratory for the analysis of aircraft ground trajectories at a given airport using ADS-B data. This tool can automatically genera... 展开 This paper presents the results of the development and validation of a tool created at the LARCASE laboratory for the analysis of aircraft ground trajectories at a given airport using ADS-B data. This tool can automatically generate a graph structure of any airport using public data available on OpenStreetMap. In addition, attributes are also defined to provide specifications for each segment defining the graph, such as segment type, segment name, bearing, distance, and speed limit. The tool also includes a map-matching algorithm that allows users to determine aircraft positions at an airport or reconstruct trajectories for statistical analysis. The map matching algorithm is based on a Hidden Markov Model process and can be used "on-line" or "off-line". An analysis of over 70 simulated trajectories showed that the algorithm was accurate to within 97-99%. The results also showed that the algorithm was able to process over 100 trajectory data points in less than 2 seconds, which is very fast. Finally, the algorithm was also tested using ADS-B data collected from FlightRadar24.com, and the results obtained were very good. 收起